Hostname: page-component-586b7cd67f-g8jcs Total loading time: 0 Render date: 2024-11-30T20:09:28.044Z Has data issue: false hasContentIssue false
  • English
  • Français

Political Cycles and Exchange Rate-Based Stabilization

Published online by Cambridge University Press:  13 June 2011

Hector E. Schamis  and
Christopher R. Way
Hector E. Schamis
Affiliation:
School of International Service, American University, Washington, D.C
Christopher R. Way
Affiliation:
Cornell University
Get access

Abstract

Fixed exchange rates implemented to increase credibility and attract foreign investment, or as an alternative approach to stabilization in high inflation economies, have been shown to be successful in expanding economic activity and correcting inflation in the short term but often lead to undesirable outcomes in the medium to long term. Although there is a wealth of literature documenting the boom-bust cycles associated with the adoption of a nominal anchor, no adequate explanation has been proffered as to governments repeatedly choosing policies that are self-defeating. The authors address this question with a political economy explanation based on the notion of a self-interested government for which short-term stabilization is attractive in the face of incentives posed by the electoral cycle. If the election coincides with the boom phase of an exchange rate—based stabilization, incumbent governments increase the likelihood of gaining reelection, and the bust phase will develop, if at all, only after the contest. The authors thus suggest a modified version of the traditional political business cycle. From the standpoint of a self-interested office-seeking government, exchange rate—based stabilizations can make good political sense. Empirical results based on an analysis of eighteen Latin American countries from 1970 to 1999 support the insights of the argument: movements towards a more fixed exchange rate are greater than three times more likely in a preelection period than in other stages of the electoral cycle.

Type
Research Article
Information
World Politics , Volume 56 , Issue 1 , October 2003 , pp. 43 - 78
Copyright
Copyright © Trustees of Princeton University 2003

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 This happened in the OECD countries, for example, during the ERM crisis of September 1992. See Buiter, Willem, Corsetti, Giancarlo, and Roubini, Nouriel, “Excessive Deficits: Sense and Nonsense in the Treaty of Maastricht,” Economic Policy 16 (April 1993)Google Scholar; and Eichengreen, Barry and Wyplosz, Charles, “The Unstable EMS,” Brookings Papers on Economic Activity 1 (1993)Google Scholar.

2 For example, Calvo, Guillermo, “Temporary Stabilization: Predetermined Exchange Rates,” Journal of Political Economy 94 (December 1986)CrossRefGoogle Scholar; Calvo, Guillermo and Végh, Carlos, “Exchange-Rate- Based Stabilization under Imperfect Credibility,” International Monetary Fund Working Papers 91/77 (1991)Google Scholar; Kiguel, Miguel and Liviatan, Nissan, “The Business Cycle Associated with Exchange Rate- Based Stabilizations,” World Bank Economic Review 6 (May 1992)CrossRefGoogle Scholar; Hoffmaister, Alexander and Végh, Carlos, “Disinflation and the Recession-Now-Versus-Recession-Later Hypothesis: Evidence from Uruguay,” International Monetary Fund Working Papers 95/99 (1995)Google Scholar; Tornell, Aaron and Velasco, Andrés, “Fixed versus Flexible Exchange Rates: Which Provides More Fiscal Discipline?” Journal of Monetary Economics AS (April 2000)CrossRefGoogle Scholar.

3 Krugman, Paul, “A Model of Balance-of-Payments Crises,” Journal of Money, Credit, and Banking 11 (August 1979)CrossRefGoogle Scholar.

4 As espoused by Nordhaus, William, “The Political Business Cycle,” Review of Economic Studies 42 (April 1975)CrossRefGoogle Scholar; and Tufte, Edward, Political Control of the Economy (Princeton: Princeton University Press, 1978)Google Scholar. Later contributions along these lines have substituted fully rational voters for the myopic electorate of the original versions, relying instead on voters' imperfect information about government competence to drive the models. See, for example, Cukierman, Alex and Meltzer, Alan, “A Positive Theory of Discretionary Policy, the Cost of Democratic Government, and the Benefits of a Constitution,” Economic Inquiry 24 (July 1986)CrossRefGoogle Scholar; and Rogoff, Kenneth and Sibert, Anne, “Elections and Macroeconomic Policy Cycles,” Review of Economic Studies 55 (January 1988)CrossRefGoogle Scholar.

5 See Mundell, Robert, “A Theory of Optimal Currency Areas,” American Economic Review 51 (September 1961)Google Scholar; McKinnon, Ronald, “Optimum Currency Areas,” American Economic Review 53 (September 1963)Google Scholar; Tavlas, George S., “The Theory of Monetary Integration,” Open Economies Review 5 (April 1994)CrossRefGoogle Scholar.

6 Giavazzi, Francesco and Pagano, Marco, “The Advantage of Tying One's Hands: EMS Discipline and Central Bank Credibility,” European Economic Review 32 (June 1988)CrossRefGoogle Scholar.

7 Cohen, Benjamin J., “Phoenix Risen: The Resurrection of Global Finance,” World Politics 48 (January 1996)CrossRefGoogle Scholar; Jeffry Frieden and Ronald Rogowski, “The Impact of the International Economy on National Policies: An Analytical Overview,” and Haggard, Stephan and Maxfield, Sylvia, “The Political Economy of Financial Liberalization in the Developing World,” both in Keohane, Robert and Milner, Helen, eds., Internationalization and Domestic Politics (Cambridge: Cambridge University Press, 1996)Google Scholar; and Strange, Susan, Mad Money (Ann Arbor: University of Michigan Press, 1998)CrossRefGoogle Scholar.

8 For the former, see Frieden, Jeffry, “Exchange Rate Politics: Contemporary Lessons from American History,” Review of International Political Economy 1 (Spring 1994)CrossRefGoogle Scholar; Geoffrey Garrett and Peter Lange, “Internationalization, Institutions, and Political Change,” in Keohane and Milner (fn. 7). For the latter, see Frieden, Jeffry, “The Politics of Exchange Rates,” in Edwards, Sebastian and Nairn, Moises, eds., Mexico 1994: Anatomy of an Emerging Market Crash (Washington, D.C.: Carnegie Endowment for International Peace, 1997)Google Scholar; Diaz-Bonilla, Eugenio and Schamis, Hector E., “From Redistribution to Stability: The Evolution of Exchange Rate Policies in Argentina, 1950–98,” in Frieden, Jeffry and Stein, Ernesto, eds., The Currency Game: Exchange Rate Politics in Latin America (Baltimore: Johns Hopkins University Press, 2001)Google Scholar.

9 Simmons, Beth, Who Adjusts? Domestic Sources of Foreign Economic Policy during the Interwar Years (Princeton: Princeton University Press, 1994)Google Scholar.

10 Bernhard, William and Leblang, David, “Democratic Institutions and Exchange Rate Commitments,” International Organization 53 (Winter 1999)CrossRefGoogle Scholar.

11 Leblang, David, “Domestic Political Institutions and Exchange Rate Commitments in the Developing World,” International Studies Quarterly 43 (December 1999)CrossRefGoogle Scholar.

12 The rendition of the dynamics of a typical exchange-rate-based stabilization is based on Kiguel and Liviatan (fn. 2); Calvo, Guillermo and Végh, Carlos, “Inflation Stabilization and Nominal Anchors,” Contemporary Economic Policy 12 (April 1994)CrossRefGoogle Scholar; and idem, “Inflation Stabilization and BOP Crises in Developing Countries,” in Taylor, John B. and Woodford, Michael, eds., Handbook of Macroeconomics (New York: Elsevier, 1999)Google Scholar. Stanley Fisher advances macroeconomic models to document the strikingly different trajectories of exchange rate-based versus money-based stabilizations; Fisher, , “Exchange Rate versus Money Targets in Disinflation,” in Fisher, , ed., Indexing, Inflation, and Economic Policy (Cambridge: MIT Press, 1986)Google Scholar. Calvo and Végh and Fisher, Ratra, and Végh survey the various theoretical models that have been put forth to account for the empirical regularities of ERBS; Calvo and Végh, “Inflation Stabilization and BOP Crises in Developing Countries,” and Fisher, Stanley, Ratra, Shay, and Végh, Carlos, “Modern Hyper- and High Inflations,” both in Journal of ‘Economic Literature 40 (September 2002)Google Scholar.

13 A variety of macroeconomic models have been proposed to account for the real effects of ERBS. A partial list includes Calvo and Végh (fn. 2); De Gregorio, Jose, Guidotti, Pablo, and Végh, Carlos, “Inflation Stabilization and Consumption of Durable Goods,” Economic Journal 108 (January 1998)CrossRefGoogle Scholar; Fisher (fn. 12); Lahiri, Amartya, “Disinflation Programs under Policy Uncertainty,” Journal of International Economics 50 (April 2000)CrossRefGoogle Scholar; Rebelo, Sergio and Végh, Carlos, “Real Effects of Exchange Rate-Based Stabilizations: An Analysis of Competing Theories,” NBER Macroeconomics Annual (1995)CrossRefGoogle Scholar.

14 Though conducive to inconsistent fiscal policies, not all exchange rate-based stabilizations are necessarily accompanied by such policies, and thus not all end with a bust in the medium run. For example, Mexico's Pacto of 1987, initiated nine months before the presidential election, was accompanied by fiscal adjustment. The downward inflation trend continued after the election, and the regime endured for several years. Similarly, Argentina's currency board arrangement appeared for quite a while to provide enduring stability, but its collapse in 2002—as well as the Mexican Tequila crisis of1994–95—cast doubt even on these “success” stories. As Michael Klein and Nancy Marion observe, while there is large variability in the duration of pegs resulting from stabilizations, few escape an eventual dramatic collapse; Klein, and Marion, , “Explaining the Duration of Exchange Rate Pegs,” NBER Working Papers 4651 (1994)Google Scholar.

15 Kiguel and Liviatan (fn. 2); Calvo and Végh (fn. 12).

16 For example, Easterly, William, “When Is Stabilization Expansionary? Evidence from High Inflation,” Economic Policy 22 (April 1996)Google Scholar; and Hamman, Javier, “Exchange-Rate-Based Stabilization: A Critical Look at the Stylized Facts,” IMF Staff Papers 48 (December 2001)Google Scholar.

17 Hamman (fn. 16), 134. Similarly, Easterly (fn. 16) finds that “consumption growth booms a bit more than output growth in the early years of stabilization” (p. 84).

18 And they are probably not wrong, as Fisher, Ratra, and Végh (fn. 12) show: “For ERBS, GDP growth rises very sharply upon stabilization and then stays high until T+2 only to fall sharply in T+3,” in “sharp contrast to non-ERBS stabilizations” (p. 868). They also reaffirm the distinctive consumption boom pattern characteristic of ERBS.

19 The evidence presented in Hamman (fn. 16), for example, suggests that the ERBS syndrome is most pronounced in Latin America, and Easterly (fn. 16) does not take issue with the stylized facts for Latin America itself, aiming only “to examine whether such stylized facts underlie the patterns detected here in a broader sample of inflation stabilizations” (p. 89).

20 Stein, Ernesto and Streb, Jorge, “Political Stabilization Cycles in High-Inflation Economies,” Journal of Development Economics 56 (June 1998)CrossRefGoogle Scholar.

21 Heymann, Daniel, “From Sharp Disinflation to Hyperinflation, Twice: The Argentine Experience,” in Bruno, Michael, ed., Lessons of Economic Stabilization and Its Aftermath (Cambridge: MIT Press, 1991)Google Scholar.

22 Nordhaus (fn. 4); and Tufte (fn. 4).

23 Cukierman and Melzer (fn. 4); Rogoff and Sibert (fn. 4); and Persson, Torsten and Tabellini, Guido, Macroeconomic Policy, Credibility, and Politics (New York: Harwood Academic Publishers, 1990)Google Scholar.

24 Edwards, Sebastian, “The Political Economy of Inflation and Stabilization in Developing Countries,” Economic Development and Cultural Change 42 (January 1994)CrossRefGoogle Scholar.

25 More precisely, in a separating equilibrium competent governments successfully signal their type by achieving stabilization, whereas in a pooling equilibrium, incompetent governments mimic competent ones in attempting stabilization. The main difference from standard competence models of political business cycles is, as shown by Persson and Tabellini (fn. 23), that the signal is inflation rather than output; voters observe inflation, but not debt, before the election. Along these lines, Stein and Streb (fn. 20) model stabilization policies in a competence framework, although they do not test their model against data. F. van der Ploeg establishes that stabilizations can have expansionary real effects even with rational, forward-looking agents able to anticipate future economic events; van der Ploeg, , “Optimal Government Policy in a Small Open Economy with Rational Expectations,” International Economic Review 28 (June 1987)Google Scholar. We thank John Freeman for bringing these arguments to our attention.

26 On competence models, see Stein and Streb (fn. 20). Economists have developed a variety of theoretical models producing the real effects associated with ERBS; for a recent survey, see Fisher, Ratra, and Végh (fn. 12). Fisher (fn. 12) and Carlos Rodriguez have focused on sticky prices in the real economy; see Rodriguez, , “The Argentine Stabilization of December 20th,” World Development 10 (September 1982)CrossRefGoogle Scholar. De Gregorio, Guidotti, and Végh have focused (fn. 13) on intertemporal shifts in durable goods purchase arising from consumers who follow inventory-type rules; Calvo (fn. 2) and Calvo and Végh (fn. 2) focus on intertemporal shifts in consumption resulting from a lack of policy credibility; and Rebelo and Végh (fn. 13) and Lahiri (fn. 13) on supply-side responses unleashed by distortions removed by stabilization. While the jury is still very much out on which of these best accounts for the data, models emphasizing intertemporal transfers in consumption and sticky prices are both compatible with our favored interpretation and enjoy some empirical support.

27 De Gregorio, Guidotti, and Végh (fn. 13) present evidence that large increases in the consumption of durables drive the boom phase of the cycle, a conclusion reaffirmed even by recent studies skeptical of many of the stylized facts about ERBS, for example, by Hamman (fn. 16). This finding suggests that ERBS is particularly attractive in societies where the majority of the population is historically excluded from access to this type of goods.

28 See Broz, J. Lawrence, “Political System Transparency and Monetary Commitment Regimes,” International Organization 56 (March 2002)CrossRefGoogle Scholar; and Frankel, Jeffrey, Schmukler, Sergio, and Serven, Luis, “Verfiability and the Vanishing Intermediate Exchange-Rate Regime,” NBER Working Papers 7901 (2000)Google Scholar. We thank an anonymous reviewer for calling our attention to this issue.

29 The eighteen countries are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Guyana, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay, and Venezuela. This is the entire population of Latin American countries with a population of at least one million for which data on exchange rate regimes, elections, and our principle control variables are readily available.

30 They could have adopted a currency board or dollarized the economy, but with the exception of Panama these were not commonly discussed policy options until recent years.

31 Under a currency board, the central bank is required to maintain liquid international reserves equivalent to (almost) 100 percent of the monetary base. As a consequence, monetary policy is directed solely at securing the exchange rate, abandoning (for the most part) the possibility of adjusting the monetary aggregate for macroeconomic stabilization purposes.

32 The descriptions that appear in the country reports of the IMF volumes provide information about timing of policy changes and allow a consistent coding (or reclassification of categories) of arrangements, neither of which is available from the summary tables in the appendixes. Besides lacking information about timing of changes, the classifications in those tables change from year to year and the categories are not defined the same way each year. In addition, the IMF yearbook data fail to capture several prominent episodes, forcing us to supplement the data with country-specific resources in some cases. We thank Jeffry Frieden for sharing his data on exchange rate arrangements in Latin America with us.

33 See Jeffry Frieden, Piero Ghezzi, and Ernesto Stein, “Politics and Exchange Rates: A Cross- Country Approach,” in Frieden and Stein (fn. 8).

34 We also have monthly data on exchange rate arrangements; in the analysis, however, we employ quarterly data because most economic data of interest are available at quarterly, but seldom monthly, intervals.

35 Our primary analyses are restricted to countries holding reasonably competitive elections and thus exclude cases of military dictatorships (and the associated plebiscites and constitutuent assembly votes). We employ these only in robustness checks.

36 Uruguay had the fewest elections, with five.

37 Although some of these elections were characterized by corruption, we include all contests, since any election can be presumed to increase the insecurity of the government, even if it has recourse to extralegal means to increase the chances of a favorable outcome. Moreover, to the extent that corrupt elections represent a negligible threat to the government, including them should make it more difficult to detect the hypothesized relationship and thus works against, not in favor of, the likelihood of finding support for our argument.

38 As shown by Kiguel and Liviatan (fn. 2).

39 Alesina, Alberto, Roubini, Nouriel, and Cohen, Gerald, Political Cycles andthe Macroeconomy (Cambridge: MIT Press, 1997)Google Scholar.

40 This is not to say that they featured no changes in exchange rate arrangements. Bolivia and Paraguay all moved unidirectionally toward more flexible rates, sometimes in several incremental steps; of the eighteen countries, only Panama saw no change whatsoever.

41 Kiguel and Liviatan (fn. 2).

42 More precisely, the bars should be of equal height until about forty-eight months and slightly shorter thereafter, since countries with short election cycles drop out at that point. While it is tempting to explore relative election cycles to avoid this shortcoming, that measure faces even graver difficulties for our purposes: the economic dynamics of ERBS are not proportional to electoral-cycle length. Because of these economic dynamics, it is precisely the calendar time until an election that is relevant, not the relative stage of the electoral calendar.

43 Using longer definitions of preelection period produces similar results. Counting the eighteen months prior to a poll, the five-category indicator produces frequencies of 4.3 percent (preelectoral) versus 2.3 percent; with the two-category indicator, the figures are 3.6 percent versus 1.6 percent.

44 Bosco, Luigi, “Determinants of the Exchange Rate Regime in LDCs: Some Empirical Evidence,” Economic Notes 16 (January 1987)Google Scholar; and Sawides, Andreas, “Real Exchange Rate Variability and the Choice of Exchange Rate Regime by Developing Countries,” Journal of International Money and Finance 9 (December 1990)Google Scholar.

45 Purchasing power parity GDP data are taken from Heston, Alan, Summers, Robert, and Aten, Bettina, Penn World Table Version 6.1 (Philadelphia: Center for International Comparisons at the University of Pennsylvania, 2002)Google Scholar.

46 Mundell (fn. 5); and Fleming, Marcus J., “Domestic Financial Policies under Fixed and Floating Exchange Rates,” IMF Staff Papers 9 (November 1962)Google Scholar.

47 Leblang, David, “Domestic and Systemic Determinants of Capital Controls in the Developed and Developing World,” International Studies Quarterly 41 (September 1997)CrossRefGoogle Scholar.

48 The only annual time-series data available for capital controls, the IMF's dichotomous coding, record merely the existence of some capital controls, rather than the severity or type of controls. By not distinguishing between intensity or extent of controls, the IMF coding is frequently misleading. For example, Mexico and Brazil are coded as equivalent in 1992, despite the fact that Brazil maintained a wide array of restrictions whereas Mexico allowed virtually unimpeded mobility. Unfortunately, better indicators are available for only a handful of years or time periods, for example, in Quinn, Dennis, “Correlates of Changes in International Financial Regulation,” American Political Science Review 91 (September 1997)CrossRefGoogle Scholar.

49 Data on reserves and M2 are taken from the IMF's International Financial Statistics CDROM. Here we follow the specification of Frieden and Stein (fn. 8).

50 Data are from the IMF (fn. 49). We also note that countries have other ways of neutralizing the volatility of a float—for example, by controls on the capital account (as in Chile) or with large reserves— and that this may obviate any relationship between financial depth and propensity to float.

51 Frieden, Ghezzi, and Stein (fn. 33).

52 Leblang(fn. 11).

53 Broz(fn.28).

54 The political freedom index emphasizes factors facilitating free and meaningful elections and is thus well suited to the focus in this article on electoral cycles. As an alternative we substituted the democracy index from the Polity IV dataset and also tried a dichotomous coding of the Freedom House index in which a country was considered electorally competitive if it received a score of 4 or higher. Both alternatives produced results very similar to those obtained with the political freedoms index.

55 Cukierman, Alex, Central Bank Strategy, Credibility, and Independence: Theory and Evidence (Cambridge: MIT Press, 1992)Google Scholar.

56 We follow the lead of Collins, Susan, “On Becoming More Flexible: Exchange Rate Regimes in Latin America and the Caribbean,” Journal of Development Economics 51 (October 1996)CrossRefGoogle Scholar; Collins includes a time trend to capture changes in the “climate of ideas” about appropriate exchange rate arrangements.

57 We thank James Vreeland for providing us with an updated version of the data on IMF agreements, expanding upon Vreeland, The IMF and Economic Development (Cambridge: Cambridge University Press, 2003)Google Scholar.

58 For an introduction to logistic regression models, see Menard, Scott, Applied Logistic Regression Analysis, Sage University Paper Series on Quantitative Applications in the Social Sciences 07–106 (1995)Google Scholar. For an intermediate textbook treatment, see Greene, William, Econometric Analysis, 4th ed. (Saddle River, N.J.: Prentice Hall, 1999)Google Scholar; and for an advanced presentation, see Maddala, G. S., Limited Dependent and Qualitative Variables in Econometrics (New York: Cambridge University Press, 1983CrossRefGoogle Scholar.

59 Beck, Nathaniel, Katz, Jonathan N., and Tucker, Richard, “TakingTime Seriously: Time-Series- Cross-Section Analysis with a Binary Dependent Vnn&ble” American Journal of Political Science 42 (October 1998)CrossRefGoogle Scholar.

60 Since changes in exchange rate regimes are infrequent, we also estimated rare events logistic regressions for each model, following King, Gary and Zeng, Langche, “Explaining Rare Events in International Relations,” International Organization 55 (Summer 2001)CrossRefGoogle Scholar; and Tomz, Michael, King, Gary, and Zeng, Langche, “RELOGIT: Rare Events Logistic Regression 1.1.”(Cambridge: Harvard University, 1999)Google Scholar. In each case the results were nearly identical to those reported here (slightly higher significance levels for U.S. real interest rates providing one exception).

61 We relax the restriction and include all countries in robustness checks described below. Using 4 as a cutoff point effectively eliminates the military dictatorships in the sample, leaving only countries holding elections featuring at least a modicum of competitiveness. The cutoff point is the same as that used by Frieden and Stein (fn. 8) to create their dummy variable for military dictatorships.

62 We report the four-quarter electoral cycle variable because it yields the strongest results. In other models, variables counting periods from nine to eighteen months as preelectoral all proved significant at the 5 percent level; significance erodes when preelection periods are defined as less than eight months or more than nineteen.

63 Running separate variables for presidential and legislative elections reveals that both produce the electoral effect, although the variable for presidential elections produces stronger results (with a coefficient of 1.4 and p-value of .002, versus 0.90 and p = .039 for legislative elections).

64 We did not include central bank independence in the models reported in Tables 4 and 5 because doing so causes us to lose two countries due to missing data. When it is included, the coefficient on CB1 is positive for both variants of the dependent variable but fails to attain significance by standard criteria (p-values of .943 and .321).

65 For example, Frieden, Ghezzi, and Stein (fn. 33).

66 Ibid.

67 This is not surprising given that IMF advice on the wisdom of ERBS programs has varied over time. However, one might still expect that a more fine-grained analysis will uncover patterns; to this end, we also interacted the IMF variable with the high-inflation variable, suspecting that the IMF pushes for ERBS programs only when inflation is a truly recalcitrant problem. Despite the intuitive appeal of this argument, the interaction term does not even approach significance, and the electoral cycle coefficients are virtually unchanged. We also tried running two separate analyses, one featuring only countries “under” the IMF and one featuring those not involved with the IMF. Results for electoral cycles were similar across the sample, reinforcing the finding that although the IMF may matter, it does not alter our core finding about patterns of electoral cycle.

68 Frieben, Ghezzi, and Stein (fn. 33).

69 We do not report results using the five-category version of the dependent variable in Table 6 to economize on space; they are virtually identical to those using the two-category variable.

70 A formal significance test also supports the obvious difference between the two types of preelection periods: a test of the hypothesis that EHIGH > ELOW allows us to reject the null hypothesis with a p-value of <.001.

71 Although EMINOR > EMAJOR, this difference is not significant by standard criteria. The difference is in the right direction, but the evidence here is merely suggestive.

72 Using the Polity IV democracy index instead produces very similar results.

73 For example, Broz (fn. 28); and Leblang (fn. 11).

74 The main drawback to such impressionistic, episodic approaches to identifying episodes is that they focus on prominent, well-known programs and fail to capture policy changes that were less extensively covered in the literature or that occurred in smaller countries.

75 Calvo and Végh (fn. 12, 1999); Kiguel and Liviatan (fn. 2); and Francisco Veiga, Jose, “What Causes the Failure of Inflation Stabilization Plans?” Journal of International Money and Finance 18 (April 1999)Google Scholar.

76 We also considered rerunning our models using the data derived from Eduardo Levy Yeyati and Federico Sturzenegger, “De Facto Classification of Exchange Rate Regimes,” February 2003 (http://www.utdt.edu/~ely/Base_2002.xls, accessed October 10,2003). However, their data are annual and do not allow us to identify the timing of changes with the precision necessary for our analysis of electoral cycles. We thus draw on the data on ERBS programs derived from episodic studies by economists as an alternative robustness check.

77 Mosteller, Frederik and Tukey, John W., Data Analysis and Regression (Reading, Mass.: Addison- Wesley, 1977)Google Scholar.

78 Arvai, Z. S. and Vincze, J., “Vulnerability of Currencies: Financial Crises in the 1990s,” Ada Oeconomica 49 (1998)Google Scholar; and Kawai, Masahiro, “The East Asian Currency Crisis: Causes and Lessons,” Contemporary Economic Policy 16 (April 1998)CrossRefGoogle Scholar.

79 For a review, see Alesina, Alberto and Roubini, Nouriel, “Political Cycles in OECD Economies,” Review of Economic Studies 59 (October 1992)CrossRefGoogle Scholar.

80 Clark, William Roberts and Hallerberg, Mark, “Mobile Capital, Domestic Institutions, and Electorally Induced Monetary and Fiscal Policy,” American Political Science Review 94 (June 2000)CrossRefGoogle Scholar; Franzese, Robert J., Macroeconomic Policies of Developed Democracies (Cambridge: Cambridge University Press, 2002)CrossRefGoogle Scholar.

81 Schultz, Kenneth A., “The Politics of the Political Business Cycle,” British Journal of Political Science 25 (January 1995)CrossRefGoogle Scholar.

82 Gaubatz, Kurt Taylor, Elections and War: The Electoral Incentive in the Democratic Politics of Peace and War (Stanford, Calif.: Stanford University Press, 1999)Google Scholar.

83 Searight, Amy E. and Way, Christopher R., “Electoral Incentives and International Bargaining: U.S.-Japan Economic Disputes and Elections, 1968–1997” (Paper presented at the annual meeting of the American Political Science Association, Washington, D.C., September 2000)Google Scholar.

84 Downs, George W. and Rocke, David, Optimal Imperfection? (Princeton: Princeton University Press, 1995)Google Scholar.

85 Tversky, Amos and Kahneman, Daniel, “The Framing of Decisions and the Psychology of Choice,” Science 211 (January 1981)CrossRefGoogle Scholar; and McDermott, Rose, Risk-Taking in International Politics (Ann Arbor: University of Michigan Press, 1998)CrossRefGoogle Scholar.